Bridge monitoring method and system based on big data analysis

Abstract

The present application relates to the technical field of statistical analysis, in particular to a bridge monitoring method and system based on big data analysis, which collects multi-component monitoring sequences and compares differences in time sequence, generates a response grading baseline according to wave threshold identification of abnormalities, constructs a jump distribution map based on amplitude grading and space-time law, identifies dense inflection points and generates a consistency curve by comparison and linkage, matches the trend of pier support to determine the stable response area, merges sequences to extract statistical characteristics and mark, and generates a bridge monitoring operation state characteristic set. The present application establishes a grading baseline based on the frequency and direction of monitoring data, forms a jump inflection point system in combination with multi-dimensional time sequence, draws a structure jump map, uses linkage identification to depict component consistency, dynamically captures local stable response, integrates trend law to complete characteristic merging, constructs state characteristic layered mapping, establishes a cross-component aggregated response chain, realizes information space-time high correlation extraction, and clearly presents the structure evolution law.

Description

Technical Field

[0001] This invention relates to the field of statistical analysis technology, and in particular to a bridge monitoring method and system based on big data analysis. Background Technology

[0002] The field of statistical analysis technology encompasses methods for systematically processing and mathematically analyzing large amounts of data. It aims to achieve quantitative expression and characteristic analysis of the state or trend of a research object by collecting, organizing, modeling, and mining patterns from raw data. The core content of this technology includes statistical modeling, feature extraction, and parameter calculation. Through mathematical reasoning and data computation, it reveals the correlations and change patterns between data. The field of statistical analysis technology is widely used in engineering, medicine, economics, transportation, and other fields. Its overall research scope covers multiple stages such as data acquisition, data cleaning, statistical description, probability analysis, model building and validation, forming a fundamental technical system for objectively evaluating and accurately analyzing the operational status of complex systems.

[0003] Among them, the bridge monitoring method based on big data analysis refers to a monitoring method that uses multi-source monitoring data to dynamically analyze and statistically evaluate the structural status of bridges. It focuses on monitoring data such as vibration, strain, and temperature generated by bridges during long-term operation. Through centralized collection and analysis of large-scale data, a statistical analysis model is constructed to identify the characteristics of bridge structural changes. It mainly uses data association analysis, time series decomposition, parameter estimation, correlation testing, and trend calculation to quantify the statistical relationships between different data types, thereby completing the monitoring and status information update process of bridge operation.

[0004] Existing technologies for bridge monitoring generally rely on the overall cumulative analysis of monitoring data, lacking detailed response differentiation based on the time progression process. When faced with multi-source component data, they cannot reflect the synchronous changes between different parts. When local components experience short-term fluctuations or mutual influences, traditional statistical models struggle to accurately reveal interrelated characteristics under the intertwined conditions of time and space. This leads to an averaging trend in data expression, weakening of difference changes, and the inability to effectively identify the local response range in continuous monitoring sequences. Monitoring data in multi-regional conditions are prone to temporal overlap and feature ambiguity, resulting in a lack of hierarchy and regional correspondence in structural status judgment. The comprehensive analysis results are prone to response lag and identification bias. Summary of the Invention

[0005] To address the technical problems existing in the prior art, this invention provides a bridge monitoring method and system based on big data analysis. The technical solution is as follows: A bridge monitoring method based on big data analysis includes the following steps: S1: Obtain the strain, displacement, acceleration and temperature sequences of the main beam, pier, bridge deck and bearings, sort them by time and perform difference comparison, identify the increase and decrease and mark the direction, compare the fluctuation amplitude with the set threshold, mark the abnormal value, statistically analyze the difference frequency and distribution by component group, and generate a bridge monitoring response graded baseline set. S2: Based on the difference records and time of the main beam and bridge deck in the bridge monitoring response graded baseline set, a monitoring sequence is constructed, the levels are divided according to the difference value, the jump level is assigned, the jump inflection point is determined and summarized, and the data is organized according to the spatial correspondence to obtain the bridge structure jump distribution map. S3: Based on the temporal clustering characteristics of the jump inflection points in the bridge structure jump distribution map, analyze the changes in strain and displacement direction of the main beam and bridge deck, compare the change direction of different monitoring points, extract the linkage events and smooth them, and generate the bridge cross-measuring point response consistency curve. S4: Call the non-abrupt sections in the bridge cross-measuring point response consistency curve, match the pier and bearing monitoring sequences, extract the overall change and speed trend, determine the deviation and register the stable response area, record the corresponding spatial location number and monitoring item, and obtain the set of stable local response sections of the bridge.

[0006] As a further aspect of the present invention, the bridge monitoring response classification baseline set includes component layer identification, abnormal fluctuation parameters, directional distribution characteristics, and difference frequency characteristics; the bridge structure jump distribution map includes time distribution information, spatial distribution information, amplitude classification information, and corresponding relationship information; the bridge cross-measuring point response consistency curve includes consistency coefficient, directional correlation characteristics, linkage event distribution characteristics, and change trend parameters; and the bridge local response stable segment set includes time segment information, spatial location number, monitoring item identification, and deviation error parameters.

[0007] As a further aspect of the present invention, the step of obtaining the bridge monitoring response hierarchical baseline set is as follows: S101: Acquire the strain, displacement, acceleration and temperature monitoring sequences of the main beam, pier, bridge deck and bearings, arrange the monitoring records in the order of acquisition time, perform subtraction operation on the values ​​of monitoring records at adjacent time points, calculate the difference in value change for each type of monitoring sequence, distinguish the direction of change of the records according to the positive or negative sign of the difference, register the change direction results with time position, and generate the change direction difference sequence. S102: Based on the time position difference in the change direction difference sequence, compare the absolute value with the set fluctuation amplitude threshold, determine the time position and monitoring point number of the difference exceeding the threshold, perform a marking operation on the record that meets the condition in the original monitoring sequence, and sort the marked records into groups according to main beam, pier, bridge deck and support, obtain the number and time distribution of abnormal records under each group, and obtain the abnormal distribution characteristics. S103: Call up the data of each group in the abnormal distribution characteristics, calculate the proportion of differences in the number of occurrences of each group in adjacent monitoring periods according to the frequency and time distribution characteristics of abnormal records, divide the response level intervals according to the proportion intervals, classify and number the group monitoring data according to the level intervals, and establish a bridge monitoring response grading baseline set.

[0008] As a further aspect of the present invention, the step of obtaining the bridge structure jump distribution map is as follows: S201: Based on the difference records and time positions of the main beam and bridge deck of the bridge monitoring response grading baseline, extract the monitoring values ​​of the main beam and bridge deck at the same time and perform difference calculation. Combine the difference results with the corresponding time positions in sequence to form a difference change monitoring sequence. Perform interval division according to the range of the absolute value of the difference in the sequence, and define each interval as a different amplitude level to obtain the difference amplitude level interval. S202: Based on the hierarchical division results in the difference amplitude level interval, the time positions falling into the same interval are marked as the same jump gear. For the continuous jump gear sequence of each monitoring point number, the adjacent time interval is calculated. For records with a time interval less than the set time threshold, the relationship between the difference between adjacent gears and the offset threshold is compared. Time points where the gear difference exceeds the offset threshold are filtered and marked to obtain the jump inflection point information set. S203: Call the time identifier and spatial location data of the jump inflection point information set, perform time sorting and spatial grouping calculation on the inflection point information according to the spatial correspondence between the main beam and the bridge deck, count the spatial concentration and distribution trend of the inflection points at each time position, and organize them in the order of time progression to form a bridge structure jump distribution map.

[0009] As a further aspect of the present invention, the step of obtaining the bridge cross-measuring point response consistency curve is as follows: S301: Based on the set of transition inflection points in the bridge structure transition distribution map, extract the time identifier data of each transition inflection point, calculate the time interval between consecutive inflection points, and filter out densely distributed time periods according to the condition that the time interval is less than the set aggregation threshold. Record the start and end times of these time periods and the corresponding monitoring point numbers to establish a set of inflection point aggregation time segments. S302: Call the time information in the inflection point aggregation time segment set, perform adjacent time difference judgment on the strain sequence and displacement sequence of the main beam and bridge deck under the corresponding time segment, mark the difference greater than zero as positive change and mark the difference less than zero as negative change, compare the change direction of different monitoring points according to time in the same time segment, filter out the records with the same change direction and register them as linkage events, and obtain the linkage direction matching set; S303: Based on the time sequence information of each linkage event in the linkage direction matching set, extract the event distribution data within the same window according to the set time window, count the changes in the number of linkage events in each time window and perform sequence smoothing processing according to time continuity, depict the synchronous change trend under the time progress, and generate the bridge cross-measurement point response consistency curve.

[0010] As a further aspect of the present invention, the step of obtaining the set of stable local response sections of the bridge is as follows: S401: Call the time series in the bridge cross-measuring point response consistency curve, calculate the adjacent difference of the response values ​​of each continuous time segment of the curve, filter out the time periods with continuous differences that do not exceed the set jump threshold, and number and register the start and end times of these time periods as independent segments to generate a set of time intervals without jumps. S402: Based on the time range of the time interval set without sudden jumps, extract the monitoring sequences of piers and supports within similar time ranges, calculate the overall variation level and the trend of change rate with time of the strain, displacement and acceleration sequences within the same interval, and organize them according to the segment number. Merge the relationship between the average variation amplitude of each sequence within the interval and the time growth rate to obtain the synchronous change characteristics of piers and supports. S403: For the data on the speed trend and overall change level in the synchronous change characteristics of the bridge pier bearings, perform deviation calculation and judgment, mark the time segments with the absolute value of deviation less than the error threshold and the direction of the speed trend consistent, record the spatial location number and monitoring item name of the corresponding bridge pier and bearing, and establish a set of stable local response sections of the bridge.

[0011] As a further aspect of the present invention, the method further includes: S5: Based on the strain in each stable response zone of the bridge local response stable section, the monitoring sequences in the stable response zone are merged in chronological order, statistical and difference analysis is performed, the overall change law, fluctuation range and frequency of occurrence of jump inflection points of each monitoring item are extracted, the corresponding bridge section and component type are marked, and the bridge monitoring operation status feature set is generated in a unified archive. The bridge monitoring operation status feature set includes horizontal change features, fluctuation pattern features, inflection point frequency features, and component location correspondence features.

[0012] As a further aspect of the present invention, the step of obtaining the bridge monitoring operation status feature set is as follows: S501: Based on the number of each stable response zone in the local response stable zone set of the bridge, obtain the strain, displacement and acceleration monitoring sequence in each stable zone, merge the monitoring records in the same zone in chronological order, perform continuity check on the time interval of adjacent records, and rearrange and integrate the time-continuous and complete sequences to generate stable zone monitoring sequence. S502: Call each monitoring sequence in the stable region monitoring sequence set, perform mean, variance and range statistics on the time records of the three sequences of strain, displacement and acceleration, calculate the fluctuation range of the sequence and extract the maximum and minimum difference values, determine the frequency position of the value changes in the sequence based on the time distribution, obtain the periodic distribution characteristics of each monitoring record, and form the change law index of the monitoring item. S503: Based on the change patterns, fluctuation ranges, and inflection point time information of the monitoring project change patterns indicators, classify and mark the corresponding monitoring projects according to the bridge cross-section span location and component classification, organize and archive all stable response zone data according to time sequence and spatial location number, and establish a bridge monitoring operation status feature set.

[0013] A bridge monitoring system based on big data analytics, the system comprising: The monitoring data processing module acquires the strain, displacement, acceleration and temperature sequences of the main beam, piers, bridge deck and bearings, sorts them by time and performs difference comparison, identifies increases and decreases and marks the direction, compares the fluctuation amplitude with the set threshold, marks abnormal values, statistically analyzes the difference frequency and distribution by component group, and generates a bridge monitoring response graded baseline set. The difference analysis and jump identification module constructs a monitoring sequence based on the difference records and time between the main beam and the bridge deck in the bridge monitoring response graded baseline set. It divides the levels according to the difference values, assigns them to jump levels, determines the jump inflection points and summarizes them. It then organizes them according to spatial correspondence to obtain a bridge structure jump distribution map. The linkage event and consistency analysis module analyzes the changes in strain and displacement direction of the main beam and bridge deck based on the temporal clustering characteristics of the jump inflection points in the bridge structure jump distribution map, compares the change direction of different monitoring points, extracts linkage events and smooths them, and generates a bridge cross-measuring point response consistency curve. The stable zone determination and response analysis module calls the non-abrupt section in the bridge cross-measuring point response consistency curve, matches the pier and bearing monitoring sequence, extracts the overall change and speed trend, determines the deviation and registers the stable response zone, records the corresponding spatial location number and monitoring item, and obtains the set of stable local response sections of the bridge. The feature extraction and state analysis module, based on the strain in each stable response zone of the bridge's local response stable section, merges the monitoring sequences in the stable response zone in chronological order, performs statistical and difference analysis, extracts the overall change pattern, fluctuation range, frequency of occurrence of jump inflection points, and time distribution characteristics of each monitoring item, marks the corresponding bridge section and component type, archives and organizes them uniformly, and generates a bridge monitoring operation state feature set.

[0014] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: In this invention, a bridge monitoring response hierarchical baseline set is established by analyzing the change direction and frequency distribution among component monitoring data. Under multi-dimensional temporal correlation, a jump level system and a jump inflection point set are formed. Based on spatial correspondence, a bridge structure jump distribution map is drawn. The consistency changes among components are characterized by directional comparison and linkage identification. During the dynamic response process, stable response characteristics of local areas are identified. By combining the change rate trends and temporal aggregation patterns of different monitoring items, feature merging and difference analysis are completed, forming a hierarchical mapping of structural state feature sets. A complete response chain is established in terms of data structuring and cross-component temporal aggregation, enabling high correlation and fine-grained extraction of monitoring information in the spatiotemporal dimension, and clearly presenting the structural state change patterns. Attached Figure Description

[0015] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a flowchart illustrating the process of obtaining the bridge monitoring response hierarchical baseline set according to the present invention. Figure 3 This is a flowchart illustrating the process of obtaining the bridge structure jump distribution map of the present invention. Figure 4 This is a flowchart illustrating the process of obtaining the bridge cross-measuring point response consistency curve according to the present invention. Figure 5 This is a flowchart illustrating the process of obtaining the set of stable local response sections of a bridge according to the present invention. Figure 6 This is a flowchart illustrating the process of obtaining the bridge monitoring operation status feature set according to the present invention. Detailed Implementation

[0016] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0017] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0018] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0019] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0020] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0021] Please see Figure 1 This invention provides a technical solution: a bridge monitoring method based on big data analysis, comprising the following steps: S1: Acquire the strain, displacement, acceleration, and temperature monitoring sequences of the main beam, piers, bridge deck, and bearings of the bridge. Arrange the monitoring records in the order of acquisition time. Perform difference comparison on the monitoring records at adjacent time points to identify the increase or decrease of the recorded values. Mark the direction of change based on the increase or decrease of the recorded values ​​and associate it with the time position. Compare the fluctuation amplitude of the recorded values ​​under each direction of change with the set fluctuation amplitude threshold. Mark the time position and monitoring point where the fluctuation amplitude exceeds the threshold as anomaly. Group the marked monitoring records into four types of components: main beam, piers, bridge deck, and bearings. Based on the frequency and distribution characteristics of the differences in each group, perform response stratification and generate a bridge monitoring response grading baseline set. S2: Based on the difference records and time locations of the main beam and bridge deck in the bridge monitoring response graded baseline set, a monitoring sequence of the difference changes between the main beam and bridge deck is constructed. Several amplitude levels are divided according to the range of difference values. The times falling into the same level are grouped into the same jump level. The time series of continuous jump level changes at the same monitoring point is judged by interval. When the time interval between adjacent jump level changes is less than the set time threshold and the level difference exceeds the set offset threshold, the time is determined as the jump inflection point. All the judged jump inflection points are summarized into a jump inflection point set according to the monitoring point number, time identifier and spatial location. According to the spatial correspondence between the main beam and bridge deck, the inflection points in the jump inflection point set are arranged in chronological order. The spatial distribution pattern under the time progression is sorted out to obtain the jump distribution map of the bridge structure. S3: Based on the set of inflection points in the bridge structure jump distribution map, extract the set of inflection points and identify the time segments in which the inflection points appear densely. Perform direction discrimination on the strain and displacement monitoring sequences of the main beam and bridge deck within the time segment. Mark the difference between adjacent times greater than zero as positive change and the difference between adjacent times less than zero as negative change. Compare the change direction of different monitoring points at each time segment within the same time segment. Extract the cases where the change direction is consistent and register them as linkage events. Organize the linkage change record sequence according to the distribution of linkage events in the window within the set time window. Perform smoothing processing in chronological order and depict the change trend to generate the bridge cross-measuring point response consistency curve. S4: Call the time segment in the bridge cross-measuring point response consistency curve that changes continuously without abrupt changes. Within the time segment without abrupt changes, call the monitoring sequences of piers and bearings with similar time ranges for matching. Within the matching range, extract the overall change level and the trend of change over time of the pier and bearing monitoring records. Determine the deviation between the trend and the overall change level. When the deviation is within the set error range and the trend is in the same direction, the time segment that meets the conditions is determined and registered as a stable response area. Record the spatial location number of the pier and bearing and the name of the monitoring project corresponding to the stable response area to obtain the set of stable local response segments of the bridge. S5: Based on the strain, displacement and acceleration monitoring sequences within each stable response zone of the bridge's local response stable section, the monitoring sequence records within the stable response zone are merged in chronological order. Statistical and difference analysis is performed on the merged monitoring sequences to extract the overall horizontal change pattern, fluctuation range morphology, and frequency and time distribution characteristics of jump inflection points for each monitoring item. The extracted features are then marked one by one with the bridge cross-section span position and component type. All monitoring feature records for the stable response zones are uniformly archived and organized to generate a bridge monitoring operation status feature set.

[0022] The bridge monitoring response grading baseline set includes component layer identification, abnormal fluctuation parameters, directional distribution characteristics, and difference frequency characteristics. The bridge structure jump distribution map includes time distribution information, spatial distribution information, amplitude grading information, and corresponding relationship information. The bridge cross-monitoring point response consistency curve includes consistency coefficient, directional correlation characteristics, linkage event distribution characteristics, and change trend parameters. The bridge local response stable segment set includes time segment information, spatial location number, monitoring item identification, and deviation error parameters. The bridge monitoring operation status feature set includes horizontal change characteristics, fluctuation morphology characteristics, inflection point frequency characteristics, and component location correspondence characteristics.

[0023] Please see Figure 2 The steps for obtaining the bridge monitoring response classification baseline set are as follows: S101: Acquire the strain, displacement, acceleration and temperature monitoring sequences of the main beam, pier, bridge deck and bearings, arrange the monitoring records in the order of acquisition time, perform subtraction operation on the values ​​of monitoring records at adjacent time points, calculate the difference in value change for each type of monitoring sequence, distinguish the direction of change of the records according to the positive or negative sign of the difference, register the change direction results with time position, and generate the change direction difference sequence. Fiber optic strain sensors deployed at key sections of the bridge are used to acquire micro-strain data of the main beam and bridge deck. GNSS displacement monitoring stations are used to acquire three-dimensional displacement data of the piers. MEMS accelerometers are used to acquire vibration acceleration data of the supports. Patch thermocouples are used to acquire surface temperature data of each component. All sensors are connected to a high-frequency data acquisition instrument via an industrial bus. Each monitoring sequence is mapped to a one-dimensional array set, and array indices are defined. Indicates the collection order, sets Increasing from 1 to , To determine the total number of sampling points in a single processing cycle, iterate through the strain, displacement, acceleration, and temperature arrays corresponding to the four types of components mentioned above, and extract the indices. With index For the corresponding numerical record, perform the difference operation of subtracting the preceding term from the following term; That is, calculation ,in Representing specific physical quantity monitoring readings, for the main beam strain monitoring sequence, if the reading corresponding to index 10 is 500.05 microstrain and the reading corresponding to index 11 is 500.08 microstrain, then a subtraction operation is performed to obtain a difference of 0.03 microstrain. For the pier displacement monitoring sequence, if the reading corresponding to index 10 is 12.50 mm and the reading corresponding to index 11 is 12.48 mm, then a subtraction operation is performed to obtain a difference of -0.02 mm. The obtained differences... Perform sign bit determination, if If the value is greater than zero, then mark the position as "1" to indicate a positive increment. If it is less than zero, then mark the position as "-1" to indicate a reverse decrement. If the difference is zero, it is marked as "0". A triplet data structure containing time index, difference value and direction mark is constructed. The above difference and marking process is executed on all monitoring channels of main beam, pier, bridge deck and bearing in sequence. All calculated differences are retained to four decimal places and repackaged according to the original array index order to generate a sequence of difference values ​​with changing direction.

[0024] S102: Based on the time position difference in the change direction difference sequence, compare the absolute value with the set fluctuation amplitude threshold, determine the time position and monitoring point number of the difference exceeding the threshold, perform a marking operation on the record that meets the condition in the original monitoring sequence, and sort the marked records into groups according to main beam, pier, bridge deck and bearing, obtain the number and time distribution of abnormal records under each group, and obtain the abnormal distribution characteristics. Based on the time differences in the sequence of changes in direction of change, a set of fluctuation amplitude thresholds is defined: ; in Set to 2.0 microstrain, Set to 0.5 mm. Set to 0.1g, The threshold was set to 0.2 degrees Celsius, based on the standard deviation of historical monitoring data under no-load conditions. The calculated result is taken as follows: Using a baseline threshold, the difference sequence of changing directions is traversed, and the difference at each time index is extracted. And calculate its absolute value. ,Will The values ​​are compared with the corresponding physical quantity thresholds. For example, for a strain difference of 0.03 microstrain in the main beam, the absolute value of 0.03 is calculated and compared with the set threshold of 2.0. If 0.03 is less than 2.0, the point is marked as normal fluctuation. If the acceleration difference of the bridge deck is 0.15g at a certain moment, the absolute value of 0.15 is calculated and compared with the set threshold of 0.1g. If 0.15 is greater than 0.1, the point is considered to exceed the threshold. The array index position of the point and the unique identification code of the monitoring point are recorded. A Boolean mask array is created, and all index positions with a judgment result of "true" are set to True, and the rest are set to False. According to the prefix field of the unique identification code of the monitoring point, all records marked as True are classified into four logical grouping containers: main beam, pier, bridge deck, and support. The total number of True values ​​in each container is counted, and the indices of these abnormal points are mapped back to the original time axis to form a sparse anomaly matrix for each component category, thus obtaining the anomaly distribution characteristics.

[0025] S103: Call up the data of each group in the abnormal distribution characteristics, calculate the proportion of the number of times the difference occurs in each group in adjacent monitoring periods according to the frequency and time distribution characteristics of the abnormal records, divide the response level interval according to the proportion interval, classify and number the group monitoring data according to the level interval, and establish a bridge monitoring response graded baseline set. Retrieve data from each group in the abnormal distribution features and set a sliding statistical window of fixed length. For example, take Each sampling point is used to move a window along the time axis within each group, and the total number of records marked as outliers within that window is counted. Execution ratio calculation formula For example, in the current window of a certain main beam strain monitoring group, the number of abnormal points counted is 150; Then calculate the proportion Preset response level range lookup table, set range correspond , correspond , correspond The calculated proportion Compare with the values ​​at the endpoints of the interval; if If it is determined to be a "weak response", If it is judged as a "moderate response", The result was determined to be a "strong response." Regarding the 0.15 in the above example, a comparison showed it fell within the range of... The starting point of the interval is determined to be a "strong response." The above proportional calculation and interval mapping operations are repeated for each sliding window of the main beam, piers, bridge deck, and supports, assigning a corresponding level label to the data in each time window. This will include the component ID, window start index, and response ratio value. and rating labels The structured records are stored sequentially in the database table to establish a bridge monitoring response grading baseline set.

[0026] Please see Figure 3 The steps for obtaining the bridge structure jump distribution map are as follows: S201: Based on the difference records and time positions of the main beam and bridge deck at the bridge monitoring response grading baseline, extract the monitoring values ​​of the main beam and bridge deck at the same time and perform difference calculation. Combine the difference results with the corresponding time positions in sequence to form a difference change monitoring sequence. Perform interval division according to the range of the absolute value of the difference in the sequence and define each interval as a different amplitude level to obtain the difference amplitude level interval. Based on the difference records and time locations of the main beam and bridge deck in the bridge monitoring response grading baseline set, the baseline set database is traversed to retrieve all record rows marked as main beam and bridge deck, and the same timestamp is extracted. Monitoring values ​​of the main beam below Bridge deck monitoring values ; Perform arithmetic subtraction For example, at a certain moment the strain of the main beam is 500. The bridge deck strain is 450. The calculated difference value is 50. The calculated difference value With the corresponding timestamp Store the data in a linear list in ascending order of time to construct a difference change monitoring sequence, and count the absolute value of all difference values ​​in the sequence. Calculate its maximum value and minimum value Set the number of amplitude levels (For example ), calculate hierarchical step size Utilizing step size to build A continuous range of values ,in Take from 1 to For example, if the difference range is 0 to 100 If the step size is 10, then generate The generated numerical intervals are defined as amplitude levels. Each level is assigned a unique integer index identifier to obtain the difference magnitude level interval.

[0027] S202: Based on the hierarchical division results in the difference amplitude level interval, the time positions falling into the same interval are marked as the same jump level. For the continuous jump level sequence of each monitoring point number, the adjacent time interval is calculated. For records with a time interval less than the set time threshold, the relationship between the difference between adjacent levels and the offset threshold is compared. Time points where the difference between levels exceeds the offset threshold are filtered and marked to obtain the jump inflection point information set. Based on the hierarchical division results within the differential amplitude range, each differential value in the differential change monitoring sequence is traversed. Compare it with each amplitude level range The boundary values ​​are compared numerically, when At that time, the location of the time. Mapped to the corresponding hierarchical integer index This involves switching gears, transforming a continuous numerical sequence into a discrete gear sequence. For each monitoring point number, the corresponding gear sequence is retrieved in chronological order, and adjacent recording points are extracted. and timestamp and gear value Perform subtraction operation Calculate the time interval and set a time threshold. Three times the sensor sampling period (e.g., sampling frequency 50Hz, period 0.02s, then...) ), the calculated and Perform a comparison, if This indicates that the change occurred within a short time window, and the absolute value of the difference between adjacent gear values ​​is then calculated. Set offset threshold There are two tiers, and Perform numerical comparison, if For example, if the gear is 2 at one moment and suddenly changes to 5 at the next moment, the difference is 3, which is greater than the threshold of 2. Therefore, this change is considered a sudden event. If the minute fluctuation is ignored and not marked, the moment that meets the judgment condition is extracted. The monitoring point number and the corresponding gear change amplitude are recorded as key data items in the results list to obtain the information set of the jump inflection point.

[0028] S203: Call the time stamp and spatial location data of the jump inflection point information set, perform time sorting and spatial grouping calculation on the inflection point information according to the spatial correspondence between the main beam and the bridge deck, count the spatial concentration and distribution trend of the inflection points at each time position, and organize them in the order of time progression to form a bridge structure jump distribution map. The system retrieves time stamps and spatial location data from the transition inflection point information set, and pre-sets a spatial coordinate mapping table between the main beam and the bridge deck. This table contains monitoring point numbers and three-dimensional coordinates. The correspondence is determined by iterating through the information set of transition inflection points, retrieving the corresponding spatial coordinates from the mapping table based on the monitoring point number, and reorganizing all inflection point data into a quintuple vector. ,in For the amplitude of the jump, it is marked by time. Perform an ascending sort operation on all quintuple vectors, set the spatial grid size (e.g., divide the bridge length into 10-meter grid cells), calculate the grid index into which each inflection point coordinate falls, and count the number of inflection points contained within each grid cell at each time step. Calculate the total number of inflection points across the entire bridge at that moment. Perform division operation The spatial concentration is quantified, and the inflection point jump amplitude within each grid is calculated simultaneously. The arithmetic mean of the calculated spatial concentrations The average jump amplitude is associated with the time progression sequence and stored to construct a data matrix that reflects the dynamic evolution of jump events in spatial geometric location over time, and to form a jump distribution map of bridge structure.

[0029] Please see Figure 4 The steps for obtaining the bridge cross-measurement response consistency curve are as follows: S301: Based on the set of transition inflection points in the bridge structure transition distribution map, extract the time identifier data of each transition inflection point, calculate the time interval between consecutive inflection points, and filter out densely distributed time periods according to the condition that the time interval is less than the set aggregation threshold. Record the start and end times of these time periods and the corresponding monitoring point numbers to establish a set of inflection point aggregation time segments. Based on the set of transition inflection points in the bridge structure transition distribution map, all record rows marked as transition inflection points are retrieved from the map data structure, and time markers are parsed from them. Arrange all inflection point time markers in ascending order to generate a one-dimensional time series. Traverse the sequence and calculate the difference between the time markers of two adjacent inflection points. Set aggregation threshold The average time for a single heavy-load vehicle to cross the bridge, for example, for a simply supported beam bridge with a span of 30 meters, is set as follows: For 5 seconds, calculate each... and Perform numerical comparison, if If these two inflection points belong to the same dense event group, then continue searching for the next interval. If continuous All intervals satisfy the condition of being less than the aggregation threshold (e.g. Then, starting from the first time the condition is met... Until the last time the condition is met Locked into a closed interval As a candidate dense time period, for each selected candidate time period, query the set of monitoring point IDs that participated in the transition within that time period, and set the start and end timestamps. The corresponding ID set is encapsulated into a tuple structure and stored in a list container to establish a set of inflection point aggregation time segments.

[0030] S302: Call the time information in the inflection point aggregation time segment set, perform adjacent time difference judgment on the strain sequence and displacement sequence of the main beam and bridge deck under the corresponding time segment, mark the difference greater than zero as positive change and the difference less than zero as negative change, compare the change direction of different monitoring points according to time within the same time segment, filter out the records with the same change direction and register them as linkage events, and obtain the linkage direction matching set; Retrieve time information from the inflection point aggregation time segment set, targeting the time segment of each record in the set. The strain sequences of the main beam and bridge deck within this time range were read in parallel from the original database. With displacement sequence Perform a first-order difference operation on each sequence; That is, calculation ,in Measurements representing strain or displacement, for Perform symbolization processing, if Assignment direction identifier (Positive change), if Assignment direction identifier (Reverse change); like Assignment At each sampling time ( Construct cross-measurement point direction vectors The subscripts represent different monitoring points. A benchmark measuring point (such as the mid-span of the main beam) is selected, and its direction is indicated. Direction identifiers of other measurement points in the vector Compare them one by one, if and Then determine the measuring point At time 10:00, the reference measurement point If a linkage occurs, If the data point is not linked, it is determined to be in a non-linkage state and discarded. The total number and specific numbers of monitoring points that meet the linkage conditions at that moment are counted. All confirmed linked monitoring point IDs and corresponding times are then recorded. The common direction of change (+1 or -1) is recorded as a linkage event entry and written into a temporary database table to obtain a linkage direction matching set.

[0031] S303: Based on the time sequence information of each linkage event in the linkage direction matching set, extract the event distribution data within the same window according to the set time window, count the changes in the number of linkage events in each time window and perform sequence smoothing processing according to time continuity, depict the synchronous change trend under the time progress, and generate the bridge cross-measurement point response consistency curve. Based on the time sequence information of each linkage event in the linkage direction matching set, the sliding time window length is set. (For example, take 2 seconds), move the window along the time axis in fixed steps (for example, 0.5 seconds), at each window position Within the window, retrieve the matching set of linkage directions and count the total number of linkage event records falling within the window range. Calculate the linkage density value For example, if 10 linked events occur within a 2-second window, the density is 5 events / second. The calculated time-density sequence... Perform the five-point cubic smoothing algorithm (Savitzky-Golay smoothing) for the sequence. Density value of points Using the formula Calculate the smoothed density value In this formula, the coefficients -3, 12, and 17 are the coefficients of the cubic polynomial fitting based on the least squares principle, and the denominator 35 is the normalization factor. The calculation aims to eliminate high-frequency random noise interference. The smoothed density values ​​are connected in chronological order to construct a two-dimensional coordinate point set with time as the horizontal axis and linkage density as the vertical axis. For the missing data segments, linear interpolation is performed to complete the data and generate the bridge cross-measurement point response consistency curve.

[0032] Please see Figure 5 The steps for obtaining the set of stable local response sections of a bridge are as follows: S401: Call the time series in the bridge cross-measuring point response consistency curve, calculate the adjacent difference of the response values ​​of each continuous time segment of the curve, filter out the time periods with continuous differences that do not exceed the set jump threshold, and number and register the start and end times of these time periods as independent segments to generate a set of time intervals without jumps. By calling the time series data from the bridge's cross-measuring point response consistency curve, the curve data is discretized into a set of time-response value pairs. Traverse the collection in ascending order of time, for index Record the point at the specified location and calculate its relationship with the successor point. absolute value of the difference Set a sudden jump threshold It is 1.5 times the global standard deviation of the curve. For example, if the global standard deviation is 0.02, then set it to... The calculated result and Perform numerical comparison, if If the change between the two points is determined to be stable, the comparison continues with the next pair of adjacent points. If the change is continuous... points (e.g.) If all elements satisfy the stationarity condition, then record the starting index of this continuous sequence. With Termination Index Extract its timestamp accordingly. and For each consecutive sequence that meets the conditions, assign a unique sequence number ID, and create a sequence containing the ID, start time, and start time. End time The system generates a structure object representing the average response value within that interval, stores all generated structure objects into a list container, and generates a set of time intervals without sudden jumps.

[0033] S402: Based on the time range of the time interval set without sudden jumps, extract the monitoring sequences of bridge piers and bearings within similar time ranges, calculate the overall variation level and the trend of change rate with time of the strain, displacement and acceleration sequences within the same interval, and organize them according to the segment number. Merge the relationship between the average variation amplitude of each sequence within the interval and the time growth rate to obtain the synchronous change characteristics of bridge piers and bearings. Based on the time range of the set of time intervals without sudden jumps, iterate through each time interval in the set. Search the pier monitoring table and bearing monitoring table in the database, and filter out all strain, displacement and acceleration records that fall within the time range; For each extracted monitoring subsequence ; Calculate its arithmetic mean To characterize the overall level of change, and simultaneously construct a time vector. Linear regression fitting is performed using the least squares method, and the slope of the regression line is calculated. The formula is The slope It has a clear physical meaning, that is, it represents the rate of change of the monitored physical quantity over time (such as displacement rate in mm / s). For example, if the fitting slope of a bridge pier displacement sequence over 10 seconds is 0.05 mm / s, then this rate value is recorded, and the calculated overall level of change is used to determine the rate of change. With the trend of rate of change Associate the data according to the segment number, and construct a system that includes monitoring point ID, physical quantity type, and The eigenvectors are used to pair and combine the pier vectors and support vectors within the same section to obtain the synchronous change characteristics of the piers and supports.

[0034] S403: For the data on the speed trend and overall change level in the synchronous change characteristics of bridge pier bearings, perform deviation calculation and judgment, mark the time segments with the absolute value of deviation less than the error threshold and the direction of the speed trend is consistent, record the corresponding spatial location number of the bridge pier and bearing and the name of the monitoring project, and establish a set of stable local response sections of the bridge. Based on the data on the speed trend and overall variation level of the synchronous change characteristics of bridge pier supports, the bridge pier rate is extracted from the paired feature vectors. With support rate and the horizontal variation of bridge piers With support variation level Execute the deviation calculation formula Error prevention logic is added here. If the value approaches 0, then a direct determination is made. If the value is 0 (considered unchanged and synchronized), otherwise a division operation is performed, and an error threshold is set. The value is 0.15 (i.e., a 15% tolerance is allowed), and the calculated value is... and Perform numerical comparison, if Continue to determine the consistency of trend direction and calculate the product. ,like If the two change in the same direction, then... If both absolute values ​​exceed the noise threshold, then a directional conflict is determined. When the deviation is less than the threshold and the directional consistency condition is met simultaneously, the time segment is marked as a "stable response". The corresponding pier ID, support ID, spatial location coordinates and specific monitoring item name (such as "pier #3 - lateral displacement") are extracted to generate a record entry containing the above metadata and establish a set of stable local response segments of the bridge.

[0035] Please see Figure 6 The steps for obtaining the feature set of bridge monitoring operation status are as follows: S501: Based on the number of each stable response zone in the local response stable section of the bridge, obtain the strain, displacement and acceleration monitoring sequence in each stable zone, merge the monitoring records in the same section in chronological order, check the continuity of the time interval between adjacent records, and rearrange and integrate the time-continuous and complete sequences to generate the stable zone monitoring sequence. Based on the numbering of each stable response zone in the bridge local response stable segment set, all unique identifiers (IDs) of stable segments stored in the database are traversed, and for each ID, the corresponding time range is retrieved from the underlying data warehouse. The original monitoring data of the main beam strain, pier displacement, and support acceleration are retrieved and loaded into memory to construct three independent one-dimensional floating-point arrays, with array indices defined. This represents the time step. For each array, an ascending sort operation based on timestamps is performed to ensure data is strictly aligned according to the order of collection. A standard time interval corresponding to the sampling frequency is set. (For example, if the sampling rate is 50Hz, then) (seconds), set the allowable time jitter error threshold. for (i.e., 0.001 seconds), traverse the sorted timestamp sequence and calculate the time difference between adjacent records. ,Will Compared with standard interval Perform numerical comparison; if the following conditions are met... Determine if the records are continuous. To determine if data loss or interruption exists, for sequences with interruptions, the array is divided into several subsequences with the interruption point as the boundary, and the length of data points in each subsequence is counted. Set a minimum effective length threshold For 500 sampling points, remove those with a length less than [a certain value]. The fragmented subsequences are reassembled in their original time order, and specific padding values ​​(such as NaN) are inserted at the splicing points to mark discontinuous boundaries, ultimately forming a cleaned and continuity-verified standardized data stream, generating a stable region monitoring sequence.

[0036] S502: Call up each monitoring sequence in the stable region monitoring sequence set, perform mean, variance and range statistics on the time records of the three sequences of strain, displacement and acceleration, calculate the fluctuation range of the sequence and extract the maximum and minimum difference values, determine the frequency position of the value changes in the sequence based on the time distribution, obtain the periodic distribution characteristics of each monitoring record, and form the change law index of the monitoring item. Call each monitoring sequence in the stable region monitoring sequence set, initialize the statistical variable accumulator, and iterate through the sequence with a length of [length missing]. Monitoring sequence array Perform the summation operation And calculate the arithmetic mean. Based on the calculated mean Iterate through the array again to calculate the square of the difference between each element and the mean. Sum and then divide by Obtain the variance Simultaneously, iterate through the array to find the maximum value. and minimum value Perform subtraction operation The result was extremely poor. Construct a first-order difference array ,in ,exist Search for the maximum absolute value in the middle and minimum value To characterize the maximum jump amplitude between adjacent time points, the mean value is set as the crossover counting baseline. traversal sequence Check two adjacent points and Compared to Positional relationships, calculate the product ,like If a mean crossover occurs, the counter is reset. Add 1 to finally calculate the crossover frequency. For example, for a displacement sequence with a mean of 10 mm, if the sequence fluctuates around 10 mm and crosses the mean line 20 times, with a total duration of 5 seconds, then the frequency is 4 Hz. The calculated... It is encapsulated into a six-dimensional feature vector to form an indicator for monitoring the changing patterns of the project.

[0037] S503: Based on the change patterns, fluctuation ranges, and inflection point time information in the indicators of the monitoring items, classify and mark the corresponding monitoring items according to the bridge cross-section span location and component classification, organize and archive all stable response zone data according to time sequence and spatial location numbering, and establish a bridge monitoring operation status feature set. Based on the changing patterns, fluctuation ranges, and inflection point time information in the monitoring project's change patterns indicators, a pre-defined bridge structure information mapping table is constructed. This table defines the hierarchical relationship between spatial location numbers (e.g., "Span-02-Section-A") and component types (e.g., "MainBeam", "Pier"). All calculated six-dimensional feature vectors are traversed, and their associated original monitoring point IDs are extracted. The mapping table is then searched for the cross-sectional span location index and component classification label corresponding to that ID. A composite data structure containing "timestamp-spatial index-component type-feature vector" is created. For different monitoring projects at the same spatial location (e.g., displacement and acceleration of the same pier), their respective feature vectors are horizontally concatenated to form a multi-source feature matrix, arranged according to the time axis. All feature matrices are globally sorted. In cases where multiple spatial cross-sectional data exist at the same time, they are arranged in spatial index order (e.g., from the bridgehead to the bridge tail). The sorted structured data is written into the document storage of a non-relational database. Each document is assigned a unique retrieval hash key, which is generated by encrypting the string "date_component type_space number" using the MD5 algorithm to ensure the uniqueness of the index. This completes the digital feature profile of the entire bridge under stable response conditions and establishes a feature set for bridge monitoring operation status.

[0038] A bridge monitoring system based on big data analytics, the system comprising: The monitoring data processing module acquires the strain, displacement, acceleration and temperature sequences of the main beam, piers, bridge deck and bearings, sorts them by time and performs difference comparison, identifies increases and decreases and marks the direction, compares the fluctuation amplitude with the set threshold, marks abnormal values, statistically analyzes the difference frequency and distribution by component group, and generates a bridge monitoring response graded baseline set. The difference analysis and jump identification module constructs a monitoring sequence based on the difference records and time of the main beam and bridge deck in the bridge monitoring response graded baseline set. It divides the levels according to the difference values, classifies them into jump levels, determines the jump inflection points and summarizes them. It organizes them according to spatial correspondence to obtain the bridge structure jump distribution map. The linkage event and consistency analysis module analyzes the changes in strain and displacement direction of the main beam and bridge deck based on the temporal clustering characteristics of the jump inflection points in the bridge structure jump distribution map, compares the change direction of different monitoring points, extracts linkage events and smooths them, and generates a bridge cross-measuring point response consistency curve. The stability zone determination and response analysis module calls the non-abrupt sections in the bridge cross-measuring point response consistency curve, matches the pier and bearing monitoring sequences, extracts the overall change and speed trend, determines the deviation and registers the stable response zone, records the corresponding spatial location number and monitoring item, and obtains the set of stable local response sections of the bridge. The feature extraction and state analysis module, based on the strain in each stable response zone of the bridge's local response stable section, merges the monitoring sequences in the stable response zone in chronological order, performs statistical and difference analysis, extracts the overall change pattern, fluctuation range, frequency of occurrence of jump inflection points, and time distribution characteristics of each monitoring item, marks the corresponding bridge section and component type, archives and organizes them uniformly, and generates a bridge monitoring operation state feature set.

[0039] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A bridge monitoring method based on big data analysis, characterized in that, Includes the following steps: S1: Obtain the strain, displacement, acceleration and temperature sequences of the main beam, pier, bridge deck and bearings, sort them by time and perform difference comparison, identify the increase and decrease and mark the direction, compare the fluctuation amplitude with the set threshold, mark the abnormal value, statistically analyze the difference frequency and distribution by component group, and generate a bridge monitoring response graded baseline set. S2: Based on the difference records and time of the main beam and bridge deck in the bridge monitoring response graded baseline set, a monitoring sequence is constructed, the levels are divided according to the difference value, the jump level is assigned, the jump inflection point is determined and summarized, and the data is organized according to the spatial correspondence to obtain the bridge structure jump distribution map. S3: Based on the temporal clustering characteristics of the jump inflection points in the bridge structure jump distribution map, analyze the changes in strain and displacement direction of the main beam and bridge deck, compare the change direction of different monitoring points, extract the linkage events and smooth them, and generate the bridge cross-measuring point response consistency curve. S4: Call the non-abrupt sections in the bridge cross-measuring point response consistency curve, match the pier and bearing monitoring sequences, extract the overall change and speed trend, determine the deviation and register the stable response area, record the corresponding spatial location number and monitoring item, and obtain the set of stable local response sections of the bridge.

2. The bridge monitoring method based on big data analysis according to claim 1, characterized in that: The bridge monitoring response grading baseline set includes component layer identification, abnormal fluctuation parameters, directional distribution characteristics, and difference frequency characteristics. The bridge structure jump distribution map includes time distribution information, spatial distribution information, amplitude grading information, and corresponding relationship information. The bridge cross-measuring point response consistency curve includes consistency coefficient, directional correlation characteristics, linkage event distribution characteristics, and change trend parameters. The bridge local response stable segment set includes time segment information, spatial location number, monitoring item identification, and deviation error parameters.

3. The bridge monitoring method based on big data analysis according to claim 1, characterized in that: The steps for obtaining the bridge monitoring response hierarchical baseline set are as follows: S101: Acquire the strain, displacement, acceleration and temperature monitoring sequences of the main beam, pier, bridge deck and bearings, arrange the monitoring records in the order of acquisition time, perform subtraction operation on the values ​​of monitoring records at adjacent time points, calculate the difference in value change for each type of monitoring sequence, distinguish the direction of change of the records according to the positive or negative sign of the difference, register the change direction results with time position, and generate the change direction difference sequence. S102: Based on the time position difference in the change direction difference sequence, compare the absolute value with the set fluctuation amplitude threshold, determine the time position and monitoring point number of the difference exceeding the threshold, perform a marking operation on the record that meets the condition in the original monitoring sequence, and sort the marked records into groups according to main beam, pier, bridge deck and support, obtain the number and time distribution of abnormal records under each group, and obtain the abnormal distribution characteristics. S103: Call up the data of each group in the abnormal distribution characteristics, calculate the proportion of differences in the number of occurrences of each group in adjacent monitoring periods according to the frequency and time distribution characteristics of abnormal records, divide the response level intervals according to the proportion intervals, classify and number the group monitoring data according to the level intervals, and establish a bridge monitoring response grading baseline set.

4. The bridge monitoring method based on big data analysis according to claim 1, characterized in that: The steps for obtaining the bridge structure jump distribution map are as follows: S201: Based on the difference records and time positions of the main beam and bridge deck of the bridge monitoring response grading baseline, extract the monitoring values ​​of the main beam and bridge deck at the same time and perform difference calculation. Combine the difference results with the corresponding time positions in sequence to form a difference change monitoring sequence. Perform interval division according to the range of the absolute value of the difference in the sequence, and define each interval as a different amplitude level to obtain the difference amplitude level interval. S202: Based on the hierarchical division results in the difference amplitude level interval, the time positions falling into the same interval are marked as the same jump gear. For the continuous jump gear sequence of each monitoring point number, the adjacent time interval is calculated. For records with a time interval less than the set time threshold, the relationship between the difference between adjacent gears and the offset threshold is compared. Time points where the gear difference exceeds the offset threshold are filtered and marked to obtain the jump inflection point information set. S203: Call the time identifier and spatial location data of the jump inflection point information set, perform time sorting and spatial grouping calculation on the inflection point information according to the spatial correspondence between the main beam and the bridge deck, count the spatial concentration and distribution trend of the inflection points at each time position, and organize them in the order of time progression to form a bridge structure jump distribution map.

5. The bridge monitoring method based on big data analysis according to claim 1, characterized in that: The steps for obtaining the bridge cross-measuring point response consistency curve are as follows: S301: Based on the set of transition inflection points in the bridge structure transition distribution map, extract the time identifier data of each transition inflection point, calculate the time interval between consecutive inflection points, and filter out densely distributed time periods according to the condition that the time interval is less than the set aggregation threshold. Record the start and end times of these time periods and the corresponding monitoring point numbers to establish a set of inflection point aggregation time segments. S302: Call the time information in the inflection point aggregation time segment set, perform adjacent time difference judgment on the strain sequence and displacement sequence of the main beam and bridge deck under the corresponding time segment, mark the difference greater than zero as positive change and mark the difference less than zero as negative change, compare the change direction of different monitoring points according to time in the same time segment, filter out the records with the same change direction and register them as linkage events, and obtain the linkage direction matching set; S303: Based on the time sequence information of each linkage event in the linkage direction matching set, extract the event distribution data within the same window according to the set time window, count the changes in the number of linkage events in each time window and perform sequence smoothing processing according to time continuity, depict the synchronous change trend under the time progress, and generate the bridge cross-measurement point response consistency curve.

6. The bridge monitoring method based on big data analysis according to claim 1, characterized in that: The steps for obtaining the set of stable local response sections of the bridge are as follows: S401: Call the time series in the bridge cross-measuring point response consistency curve, calculate the adjacent difference of the response values ​​of each continuous time segment of the curve, filter out the time periods with continuous differences that do not exceed the set jump threshold, and number and register the start and end times of these time periods as independent segments to generate a set of time intervals without jumps. S402: Based on the time range of the time interval set without sudden jumps, extract the monitoring sequences of piers and supports within similar time ranges, calculate the overall variation level and the trend of change rate with time of the strain, displacement and acceleration sequences within the same interval, and organize them according to the segment number. Merge the relationship between the average variation amplitude of each sequence within the interval and the time growth rate to obtain the synchronous change characteristics of piers and supports. S403: For the data on the speed trend and overall change level in the synchronous change characteristics of the bridge pier bearings, perform deviation calculation and judgment, mark the time segments with the absolute value of deviation less than the error threshold and the direction of the speed trend consistent, record the spatial location number and monitoring item name of the corresponding bridge pier and bearing, and establish a set of stable local response sections of the bridge.

7. The bridge monitoring method based on big data analysis according to claim 1, characterized in that: The method further includes: S5: Based on the strain in each stable response zone of the bridge local response stable section, the monitoring sequences in the stable response zone are merged in chronological order, statistical and difference analysis is performed, the overall change law, fluctuation range and frequency of occurrence of jump inflection points of each monitoring item are extracted, the corresponding bridge section and component type are marked, and the bridge monitoring operation status feature set is generated in a unified archive. The bridge monitoring operation status feature set includes horizontal change features, fluctuation pattern features, inflection point frequency features, and component location correspondence features.

8. The bridge monitoring method based on big data analysis according to claim 7, characterized in that: The steps for obtaining the bridge monitoring operation status feature set are as follows: S501: Based on the number of each stable response zone in the local response stable zone set of the bridge, obtain the strain, displacement and acceleration monitoring sequence in each stable zone, merge the monitoring records in the same zone in chronological order, perform continuity check on the time interval of adjacent records, and rearrange and integrate the time-continuous and complete sequences to generate stable zone monitoring sequence. S502: Call each monitoring sequence in the stable region monitoring sequence set, perform mean, variance and range statistics on the time records of the three sequences of strain, displacement and acceleration, calculate the fluctuation range of the sequence and extract the maximum and minimum difference values, determine the frequency position of the value changes in the sequence based on the time distribution, obtain the periodic distribution characteristics of each monitoring record, and form the change law index of the monitoring item. S503: Based on the change patterns, fluctuation ranges, and inflection point time information of the monitoring project change patterns indicators, classify and mark the corresponding monitoring projects according to the bridge cross-section span location and component classification, organize and archive all stable response zone data according to time sequence and spatial location number, and establish a bridge monitoring operation status feature set.

9. A bridge monitoring system based on big data analysis, characterized in that, The system is used in the bridge monitoring method based on big data analysis as described in any one of claims 1-8, and the system comprises: The monitoring data processing module acquires the strain, displacement, acceleration and temperature sequences of the main beam, piers, bridge deck and bearings, sorts them by time and performs difference comparison, identifies increases and decreases and marks the direction, compares the fluctuation amplitude with the set threshold, marks abnormal values, statistically analyzes the difference frequency and distribution by component group, and generates a bridge monitoring response graded baseline set. The difference analysis and jump identification module constructs a monitoring sequence based on the difference records and time between the main beam and the bridge deck in the bridge monitoring response graded baseline set. It divides the levels according to the difference values, assigns them to jump levels, determines the jump inflection points and summarizes them. It then organizes them according to spatial correspondence to obtain a bridge structure jump distribution map. The linkage event and consistency analysis module analyzes the changes in strain and displacement direction of the main beam and bridge deck based on the temporal clustering characteristics of the jump inflection points in the bridge structure jump distribution map, compares the change direction of different monitoring points, extracts linkage events and smooths them, and generates a bridge cross-measuring point response consistency curve. The stable zone determination and response analysis module calls the non-abrupt section in the bridge cross-measuring point response consistency curve, matches the pier and bearing monitoring sequence, extracts the overall change and speed trend, determines the deviation and registers the stable response zone, records the corresponding spatial location number and monitoring item, and obtains the set of stable local response sections of the bridge. The feature extraction and state analysis module, based on the strain in each stable response zone of the bridge's local response stable section, merges the monitoring sequences in the stable response zone in chronological order, performs statistical and difference analysis, extracts the overall change pattern, fluctuation range, frequency of occurrence of jump inflection points, and time distribution characteristics of each monitoring item, marks the corresponding bridge section and component type, archives and organizes them uniformly, and generates a bridge monitoring operation state feature set.

Images